Pure &Appl. Chem., Vol. 60, No. 5, pp. 753-768, 1988. Printed in Great Britain. 0 1988 IUPAC

Mechanisms of decomposition of hydrocarbons in electrical discharges

D.I. Slovetsky

Institute of Petrochemical Syntheses, Academy of Sciences of USSR, Moscow, V-71

Abstract - The mechanisms of decomposition of hydrocarbons in electric discharge plasmas are discussed, with special reference to: (i) plasma parameter measurements, (ii) composition of stable and unstable chemically active particles, (iii) mechanisms of ionization, excitation and decomposition including solid polymeric products generated by nonequilibrium chemical kinetics.

INTRODUCTION

Much work has been published on the study of the decomposition of hydrocarbons in nonequilibrium electrical discharges in which gaseous and solid polymeric products are generated, e.g. linear and cyclic saturated and unsaturated hydrocarbons and aromatic compounds. The dependence of outputs in stable products on external discharge parameters - such as pressure, gas-flow rate and electric power - have been used in the past to elucidate mechanisms of chemical reactions. The method using small accepting admixtures, applied successfully in the past for studying reactions in photo- and radiation chemistry, have not been so successful in the case of plasmas because (i) they may give a large variation of internal plasma parameters; and (ii) known radical or ion acceptors may have the opposite effect due to decomposition, ionization or excitation. Analogy of plasmachemical product composition with photochemical, radiochemical or mass-spectrometrical ones was used in interpretation also. But this way is not informative because of large variation of product composition with plasma parameters and complexity of chemical conversations with participation of electrons, radicals, excited particles and negative and positive ions.

Chemical reactions in nonequilibrium cold plasmas are initiated, as a rule, by collisions of molecules with electrons having large energy taken from electrical field immediately. Many particles having large chemical activity are generated by electron molecule collisions such as electronically and vibrationally excited molecules, negative and positive ions and radicals (ref. 1). Simultaneously some stable products may be generated. All particles may take part in following collisions with electrons, molecules, and one with another, giving secondary reactions. Heavy particles energy is low if the gas temperature is low. But by large specific power of discharge it may be large enough to give thermal pyrolysis reaction (Fig. 1,2). In addition, due to electron-molecule primary collisions hot atoms and radicals may be generated which are very active in chemical reactions even at low gas temperature.

Fig. 1. Gas temperatures on the Fig. 2. Gas temperatures at the axis of discharge through so- middle line between parallel me gases in tubular reactor plate electrodes in diod type versus specific power at I (R)= reactor versus specific power 300 K. with Le2 cm, R-L at electrodes

f=2g(0)=Tg(L)=300 K g

g

753

754 D. I. SLOVETSKY

Due to this reason the chemical reaction mechanism in nonequilibrium plasma may be very complex and dependent on internal plasma parameter range. It is necessary to note that determination of active particles, giving main contribution in product generation is difficult because of the various type of active particles there are in any electrical discharges through any gases but not all of them are important in chemistry. Hence it is necessary for the plasmachemical reaction study to use the method of choice of most probable mechanisms of chemical reactions and physical processes in nonequilibrium plasmas (ref. 1). The method is based on the quantitative study of generation and decay kinetics of any particles in the frame of nonequilibrium chemical kinetic theory (ref. To realize the method in practice is is desirable to have measurements of all internal plasma parameters which influence the direction and rate of the chemical reactions (Table

2 ) .

1) *

TABLE 'I. Discharge parameters and i n t r e n a l plasma parameters wich a r e necessary t o know f o r mechanisms studying i n nonequi- l ibr ium plasmas

_ ~ _ _ _ _ Parameters Measuring technique

Discharge parameters : Power , W , W Electrotechnical devices E l e c t r i c a l cu r ren t , I , mA E l e c t r i c a l f i e l d s t r eng th , E , V/cm Probes Gas pressure, p, Pa Membrane or capaci t ive manometers Gas flow r a t e , G , cm5/s Volume gas flow meter

I n t e r n a l plasma parameters Energy of p a r t i c l e s : Electron energy d i s t r i b u t i o n E l e c t r i c a l probes, o p t i c a l funct ion (EEDF), fe(Ee) spectroscopy

Heavy p a r t i c l e (gas) tempera- t u r e , Ig, K

Rotat ional temperature of Optical spectroscopy, l a s e r inclu- molecules , Trot , K Vibrat ional temperature or d i s t r i b u t i o n funct ion of vib- r a t i o n a l l e v e l populations

(Chermop$obes, thermocouples, Doppler broadening of s p e c t r a l l i n e s

ced, Raman spectroscopy

P a r t i c l e d e n s i t i e s : Electrons , ne E l e c t r i c a l probes, absorption or

r e f l e c t i o n of microwave r ad ia t ion , s t a r k broadening of s p e c t r a l l i n e s (ne I O ' ~ C ~ - ~ )

Ions , ni Mass-spectrometry Exciced p a r t i c l e s , ni Optical spectroscopy Radicals, nr Optical spectroscopy, mass-spect-

*

rometry, ECR-spectrometry i n gas phase, with matrix i s o l a t i o n on surfaces , r a d i c a l trapping with ECR

~ ~

Stable reagents and products Chromatography, mass-spectrometry, Chromatomass-spectrometry, I R -ab- sorpt ion, l a s e r induced f luorescent Calculation from measured pressure and gas temperature

Total heavy p a r t i c a l densiety,

NO

Mechanisms of decomposition of Hydrocarbons in electrical discharges 755

Review of the works concerning hydrocarbon chemical reactions under discharge conditions shows insufficient knowledge of plasma parameters. In many cases even external parameters, such as specific electric power, are not reported. Investigations were carried out using various reactor constructions and discharge types - direct current, high frequency and microwave discharges. Without internal plasma parameters having been measured it is difficult to compare the results of various workers and to make objective conclusions about mechanisms of chemical reactions. More detailed studies were performed using direct current discharge through mixtures of rare gases with methane and C5-C7 - hydrocarbons (ref. 3-15). parameters, radical and stable product compositions were measured. Variation of hydrocarbon contents and discharge current makes it possible to study kinetic order of reactions. Important results were obtained by means of labeled isotopic techniques which allow to discriminate the products of primary decay from products of the secondary one.

This lecture is based on the results of plasma parameter measurements and composition of stable and unstable chemically active particles. Mechanisms of ionization, excitation and hydrocarbon decomposition are discussed including generation of gaseous and solid polymeric products by means of nonequilibrium chemical kinetics.

The internal plasma

PLASMA PARAMETERS, IONIZATION AND EXCITATION MECHANISMS

Plasma parameters are best studied in positive column of direct current glow discharges through rare gases (ref. 1). But addition of small concentrations of hydrocarbons t0 rare gases gives large variation of electric field strength, mean electron energy and density (Fig. 3 ) . Most sharp variations are observed at additions of 0(=0,01-0,05% c-CbH12. At further increases of GC from 0,05 to 2%, the variations are smaller and plasma parameters are nearly constant in this range. Decreasing the electrical field strength results in hydrocarbons having ionization potential lower than excitation potential of the rare gas metastable atomic levels. When the electric field strength is increased (ref. 1,3,4,6) the result is opposite. In more typical former case, gas temperature is lowered due to high heat transfer coefficient of hydrogen generated as a decomposition product and due to the decrease in electric field and specific power. decreased as a result. Addition of hydrocarbons to rare gases decreases fast electron density even if E/No is constant (ref. 8-10). Both effects decrease coefficients of processes initiated by electron impacts and electron drift velocity. Electron density is increased to compensate decrease of drift velocity and to mintain constant electric current.

To reduce electric field, E/No is

Fig.3. Effect of addition of small concentrations of hydrocarbons to rare gases in d.c. glow discharges - on electric field strength, mean electron energy and density.

Ionization rate must increase to give larger electron density in spite of decrease of electric field. Therefore, ionization mechanisms by hydrocarbon addition are varied. Ionization proceeds in glow discharge through pure rare gas under sufficient high current density ( ) >l mA/cm2) by a stepwise path: excitation of metastable atomic levels by electron impacts followed by ionization of metastable atoms by second electron impacts (ref. 1,6). Metastable levels excitation rate is decreased due to hydrocarbon addition and metastable atom densities are lowered due to decrease of density of fast electrons and quenching by hydrocarbon molecules. At very small additions (kL0,001%) quenching is low compared to the electron impact ( 4 5 % ) . However, the electric field strength is already decreased (Fig. 3 ) . It is connected with contribution of Penning ionization occuring by collisions of metastable rare atoms with hydrocarbon molecules. The greater the hydrocarbon addition, the larger is molecular quenching and whend=l% its rate is at hundred times larger than quenching of metastables by electrons. Metastable densities are at hundred times less than the ones in rare gases. As a result the rates of stepwise processes of ionization and upper atomic levels excitation are decreased. This effect gives observable quenching of luminosity of discharge when hydrocarbons are added.

756 D. I. SLOVETSKY

Onaddition of cyclohexane t o argon more than O , O l % main ionizat ion process i s Penning ionizat ion having efficiencywhich i s equal t o 0,56+0,3 per co l l i - sion. Direct ionizat ion of hydrocarbon molecules by electron impact gives contribution only a t l a rger additions. The process i s main i n pure hydrocarbon discharges where the e l e c t r i c f i e l d s t rength i s increased again. The reason of this increase i s deficiency of f a s t electrons i n i t s energy d is t r ibu t ion Which i s the more the more are ad- di t ions. I f the ionization poten t ia l of hydrocarbons i s la rger than exci ta t ion energy of ra re gas met;astable leve l ( for example, methane has Eion=12,5 eV, and argon Em=11,55 eV) e l e c t r i c f i e l d s t rength i s increased by small e x t e n t because of decrease of stepwise ionizat ion due t o quenching of metzastable atoms which is not compensated by appearance of new ionizat ion channel ( r e f .

Penning ionizat ion was used i n ( re f . 11) t o explain the s low changiW7 of e l e c t r i c f i e l d with time i n discharge through mixtures of Ar,He and Ne w i t h benzene(&O,l%). But charac te r i s t ic time of the var ia t ion w a s equal t o t en secondswhich is much more than the survival time of metnastable atoms varying f rom i n pure ra re gases t o i n mixtures. Hence t h i s expla- nation i s not correct. It is more probable tha t s l o w var ia t ions of e l e c t r i c f i e l d as observed are forced by chemical reactions, occuring par t iculary a t reactor walls. Small var ia t ion o f plasma parameters i n mixtures of ra re gases with 0,1-V0 hydrocarbons allow t o study k ine t i c order of react ion depedendinjon hydro- carbon concentrations. Indepedence of plasma parameters on gas residence ti- me i n discharge (rp10-3s) gives poss ib i l i t y t o carry out k ine t i c studying with simple interpretat ion. Influence of by - electrode regions on the che- mical reactions was excluded by freezing of products i n by - electrode zones ( r e f . 5-7).

1,15>

MECHANISMS OF DISSOCIATION OF HYDROCARBONS

Dissociation r a t e s of cyclohexane as measured are increased according t o li- near - low w i t h increase of hydrocarbon concentration i n mixture with argon, increased monotonically depedendinaon current density and gas pressure and independent o gas residence time i n discharge within e r ro r l i m i t s (Big. 4). This independence allow t o exclude the reverse reaction of cyclohexane generation from dissociat ion products. This conclusion was confirmed by ex- periments w i t h use of mixtures of protonated and deuterated cyclohexanes (1:l and 1:lO) i n argon. I n t h i s case products d o not contain isotopic mixed cyclohexanes , It was shown by evaluation of react ion r a t e s of a l l energet ical ly possible reactions tha t the main contribution i n cyclohexane dissociat ion may give co l l i s ions with electrons, met;astable argon atoms and hot hydrogen atoms :

c-C6HI2 + e - e + neutral products , (1 1

c-C6H12 + H* - products . ( 3 ) ( 2 )

1 c-C6H12 + A r m + Ar( So) + neutral products ,

Hot hydrogen atoms (H*) may be generated due t o primary dissociat ion proces- ses (1,2) v i a decay of highly electronical ly excited molecules of cyclohexa- ne. Hot atoms d iss ipa te t h e i r energy during co l l i s ions with argon a toms be-

Mechanisms of decomposition of Hydrocarbons in electrical discharges 757

fore co l l i s ion with hydrocarbon molecules (3). Hence react ion probabi l i ty de- pends on hydrocarbon concentration i n mixture. Upper l i m i t of react ion (3) r a t e can not be more than r a t e of ho t atoms generation i n reactions (1 ,2 ) , because i n every ac t of react ion (3) hot atom must dlsappear. Therefore contribution of react ion (3) can i n be t t e r case only enlarge the dissocia- t i o n r a t e by f ac to r two. It i s the case if the every ac t of reactions (1 ,2) gives appearance of hot atom and every hot atom gives disappearance of cyc- lohexane molecule i n reaction (3). But r ea l ly contribution of this react ion i n hydrocarbon dissociat ion evidently i s not s o large becausethe dissocia- t i o n r a t e i s d i r ec t ly proportional t o cyclohexane concentration i n mixtures as observed (Fig. LCa).

Fig. 4. Dissociation r a t e s o f cyclohexane versus its addition t o argon i n mixture ( a ) , discharge current densi ty (b) and gas pres-

ces calculated: 1 - due t o reaction 1, 2 - 2; 3 - ionizat ion r a t e of cyclohexane due t o co l l i s ions with argon methastable atoms ; 4 - r a t e of so l id products generation deterinined from C -atom ba- lance. Points are dissociat ion r a t e measured

sure (c) with R=O,5 cm; G=390 cm 3 / s . Lines are r e l a t ive dependen-

T a k i n g i n to account the exci ta t ion and deexcitation of metzastable argon a t oms

Ar( 1 S o ) + e K- x4

A r m + M -Ar( Kg I So) + products

A r m + e , A r m + e 3 A r * (Ar’) + e (2e) ,

the reaction (2) r a t e equals

I n expression (7) probabi l i ty of d issociat ion due t o molecular quenching (2 , 6) is introduced

5 D = K2/K6 ( 8 ) A l s o it is taken in to account t ha t molecular qwnching is large compared w i t h ra re atom one,and cyclohexane dissociat ion degree i s small. Then using the r a t e coeff ic ients values K -(4,5+1,5) *10-”7crn3s-” ; ~~=10-9cm3s-~ ( r e f . 13,14) andd *0,5% (d is r e l a t ive concentration of cyclohexane i n i n i t i a l mixture) i t canpshown tha t react ion ( 2 ) r a t e i s not dependent on as ra- t i o K ~ ~ , / K ~ F - C ~ H ~ ~ = 5 103d LO, I and electiron density and energy are con- s t an t i n the range 0,176 s,,( I % (Fig. 3) . Rate of the react ion (3) must be

5-

758 D. I. SLOVETSKY

di rec t ly proportional t o A. A s follows gives main contribution i n the range 0,1%50(.~2%. The react ion ( 2 ) contribu- t i on i s s ignif icant only i f ~ o , I % . The same dissociation ra tes were observed f o r other C -C hydrocarbons under same conditions (Table 2 ) . Reduced e l e c t r i c f i e l d i s the same t o o f o r vario- us hydrocarbons excluding benzene.Hence poss ib i l i ty of Penning-ionization by col l is ions of metzastable argon atoms with the hydrocarbons i s nearly cons- tant . For benzenevalue of E/No i s enlarged,which connected with diminishing of Penning-ionization probabili ty due t o appearance of additional channels such as intermolecular conversion of exci ta t ion energy,

from data of Pig. 4 , react ion (3)

5 7

T A B U 2. Steps and r a t e s of decomposition of number hydrocarbons i n glow discharge through the i re mixtures with argon under the same conditions: j=2,4 mA/cm2; p=133 Pa ; kXHy =I% ; dtube=l cm ( r e f . 4,5)

~~

Hydrocarbon Legree of Residence Dissociation Reduced e l e c t r i c dissocia- time, t , r a t e , C ~ - ~ S " f i e l d , E/N , t ion , % 10-3, 10-~6V*cm 2 O

'-'6% 2 1522 3 94 1@2 192

0 12+-2 4,2 g,0?1,4 1 ,2 c-C6H1 CH3 723 394 6,5f-2,6 1 ,I

'gH1 4 1 o+-2 415 7,It1 $3 112

'SH6 1222 492 9,021,4 197

The small var ia t ion of dissociat ion r a t e s mean the approximate equa- l i t y of dissociation cross-sections under electron - hydrocarbon col l is ions. T h i s conclusion is confirmed by calculation of cross - sections of disso- c ia t ion via exci ta t ion of opt ical ly allowed electronic t rans i t ions by elec- t ron - molecular impacts (Fig. 5) ( re f . 15). The dissociation r a t e s as cal- culated using t h i s cross - sect ion and electron energy d is t r ibu t ion as cal- culated f o r pure argon plasmas ( re f . 10) are s ign i f icant ly lower than measu- red ones (Fig. 6). This discrezpancy may be explained by contribution of ex- c i t a t ion of opt ical ly inallowed electronic transitionswhich is main process f o r the hydrocarbons as was shown i n the past f o r the methane, ethane, pro- pane and butane ( re f . 16). The same dissociation mechanism i n c l u d i n g stage (1-3) was observed f o r mixture of methane with ra re gases ( re f . 6).

Fig. 5. Cross sections of hydrocarbons dissociat ion by electron impackts f o l l o - wed by exci ta t ion of opti- ca l ly allowed electronic t rans i t ions

4 0 20 30

Mechanisms of decomposition of Hydrocarbons in electrical discharges 759

Fig. 6. Rate coef f ic ien ts of proces- s e s i n i t i a t e d by e lec t ron impacts i n mixtures of argon with hydrocar- bons b(L10%) versus reduced elec- t r i c f i e l d . I -argon a tom met2astab- l e l e v e l exc i ta t ion ; 2 -d i rec t ioni- za t ion of cyclohexane molecules ; 3 -d i rec t d i ssoc ia t ion of cyclohexa- ne molecules through e x c i t a t i o n of o p t i c a l l y allowed e l e c t r o n i c transi- t ions ; 4 -the same as ( 3 ) f o r butane molecules; 5 -d i rec t d i s s o c i a t i o n of butane through e x c i t a t i o n of opti- c a l l y e l e c t r o n i c t r a n s i t i o n s ; 6 - cyclohexane d issoc ia t ion r a t e measu- red; 7 - ionizat ion r a t e measured i n discharge and determined by Penning ion iza t ion of cyclohexane

BCmZ 4 6

Thus i n i t i a t i o n of hydrocarbons decomposition under nonequilibrium plasma conditions i n mixtures with r a r e gases is f a s t e lec t rons and exci ted metzastable r a r e atoms, Direct d i s s o c i a t i o n by e lec t ron impacts ( a few t e n t h of percent) up t o 100% hydrocarbons. Radicals generated due t o primary decay do not cause 500-600 K ) . The d issoc ia t ion r a t e C ~ I I be twice due t o influence of hot a toms and r a d i c a l s generated during primary decay. The same d issoc ia t ion mecha- nisms must be observed f o r hydrocarbons i n other types of discharges such as HF and MW under the same range of parameters: Ee21,5-2 eV; xe=ne/No<10-5; T C600 K. Variation of d i ssoc ia t ion mechanism is possible due t o

t i n g a t high s p e c i f i c discharge power (Fig. 1 , 2 ) , when the r a d i c a l reac t ions may give chains. I n ( r e f . 17) it was supposed t h a t increasing dissocia- t i o n r a t e s of number hydrocarbons as observed was caused by contr ibut ion of react ions w i t h p a r t i c i p a t i o n of v ibra t iona l ly exci ted molecules without dis- cussion of gas heating e f fec ts . T h i s supposit ion seems t o be impossible be- cause of high e f f ic iency of hydrocarbon v ibra t iona l re laxa t ion ( r e f . 18).

due t o col l is ionof molecules w i t h

gives main contr ibut ion beginningfrom small additives

chain processes a t low gas temperatures (T = g

-

Q enlarged gas hea-

MECHANISMS OF GENERATION OF STABLE GASEOUS PRODUCTS

S o l i d polymeric f i l m s a re the main product of c 4 4 6 hydrocarbon decomposi- t i o n i n d i r e c t current discharge through mixtures of argon w i t h hydrocar- bons. The films are deposited on the reac tor wal ls , e lectrodes and samples inser ted t o plasma (Table 3). Gaseous products composition of C-C6H12 disso- c i a t i o n i n mixture w i t h argon i s near ly t h e same tha t one i n pure C-C6HI2 (Table 4, culomns 2,3). But change of discharge type from dc t o HF and MW discharge through methylcyclohexane gives v a r i a t i o n of gaseous product com- pos i t ion and r a t i o of outputs of gaseous and s o l i d products (Table 4, co- lumns 8-10).

760 D. I. SLOVETSKY

TABLE3. Outputs of gaseous and solid products of hydrocarbons decom- position in discharges through mixtures with argon (ref.4,5)

Hydrocarbon a( ,% Discharge Sum of gaseous Sum of solid c G Y products products

100 Constant current . - observed glow, p=133 Pa c-C6H1 2

- 11 - 14 86 2

- 1 ) 38 62 c-c5H1 0

9 91 '6'1 4

- I t - - observed 'gH6

C-C~HI 1 CH3 1 ,o 29 71 'lcHl 0

190 - 1' - 12 aa 190 1 ,o 190 190

- ' 1

- ' 1 - C-C~HI q CH3 100 High frequency(KF) 79 21 C-C6H1 CII 100 Microwave (MN) 93 7 l'ABIiE4. Main stable products of number hydrocarbons decomposition

in electrical discharges (sum of gaseous product without hydrogen i s equal to 100%)

~~ ~

Stable Constant current glow discharge,p=133 Pa KF-dis- MW-dis- product charge charge

Mechanisms of decomposition of Hydrocarbons in electrical discharges 761

Gaseous products are main i n HF and MW discharges (Table 3 ) . Products gene- rated due t o breaking o f f C-C bonds predominate among the gaseous products i n dc discharges, A t t h a t time products generated due t o breaking off C-H bonds predominate among the decomposition products i n HF and MW discharges, where more partsof products are gaseous ones. I n e a r l i e r works ( r e f , 7) it had been supposed t h a t gaseous products were generated mainly due t o decay of cyclohexane r ings prefzerably t o s tab le molecular fragments of C2:C2:C2; C * C and C2:C4 types. But it is c lear t ha t decay can give radicalswhich being unstable pa r t i c l e s

Determination of radicals i n dischage plasmas is one of the more complex problems i n studying dissociat ion mechanisms (Table I). I n e a r l i e r works radicals were ident i f ied by means of op t ica l spectra of plasmas, But this method gives cer ta in ident i f ica t ion a s ru les only f o r a toms and two atomic rad ica ls and i n exclusive cases f o r +atomic rad ica ls (ref.192 I n the cases of manyatomic radicals the ident i f ica t ion i s d i f f i c u l t due t o diffusion of spectra and hence i s based on chemical proof only. It gives uncertain conclusions about type of radiat ing rad ica l (Table 5). Mass-spec- troscopic technique gives a cer ta in information i n the case of l i g h t hydro- carbons (C1,C2) , but f o r heavier ones d i f f i c u l t i e s of ident i f ica t ion are ve- r y large. Spin trapping technique was found t o be the mos t e f f e c t i v e

hydrocarbons.Anumbefqsubstances(such as 2-methyl-2-nitrozopropane, nitroben- zene,phenyl-thret-butylnitron e tc ) exposed t o plasma can capture the radicals giving s tab le substances(so-called adduct-radica1s)with charac te r i s t ic ESR- spectra, depending on type of radicals . Exposing of sp in trapping t o plasma of discharge i n pure argon gives no ESR-signals. I n l a t t e r case surface of spin trapping positioned a t 5 cm distance from discharge i n gas flow is bom- barded by excited and charged particleswhich flow i s much more than i n mix- tu re of argon with hydrocarbons. Therefore only hydrocarbon radicals give ESR-signal (Fig. 7) . Information on type of rad ica ls i s derived from pick t o pick distance and pick in t ens i t i e s i n ESR-spectra of adduct-radicals. The view of adduct-radical ESR-spectra i s differentfrom ESR-spectra of the f r e e radicals ,

3' 3 to give r isekstable products as r e su l t of secondary reactions,

one f o r heavy

Discharge through kr+ c-C6H1

Fig. 7. ESR-spectra of adduct-radicals observed a f t e r exposition of rad ica l trapping t o discharge through mixture of C6HI2 with argon and pure argon

762 D. I. SLOVETSKY

TABLE 5. Main rad ica ls* ) , observed i n nonequilibrium e l e c t r i c discharges through hydrocarbons

Ref e- t e chnique rence

Hydrocarbon Observed r a d i c a l s Measuring Conditions

CH3,CH2,C$13,C2H5,H Mass-spec- HF-discharge, (ref.20) tr ome t r p=13-133 Pa CH4

cH4 H,CH,CH2, C2 C 3 Optical Impuls dischar- ( ref .19) spe c t r 0- ge

H Constant current SCOPY glow discharge (ref .6)

'gH6 CGH5(?) Optical Constant current o r spectro- glow discharge '6 H5 CH3

C6H5C3H7 C,H5GH2( ?) SCOPY through mixture C6H5CH2C6H5 of heleum with C6HCj(CH2) 2CGH5 I - 5% hydrocarbons

p=26-200 Pa ; j = ~ - 2 mA/cm2 (ref.21)

'gH6 H, CH, C 2 , C2H( o r C4H'2> Optical C H CH ( ? ) , C 6 H 5 C ( ? ) spectro- charges , C6H4(?), C6H5CH2(?) SCOPY p26-266 Pa ;

HF and MW-dis-

6 5 2

W = 15-100 W (ref.22)

'gH6 'gH5 Radical KF-discharge (ref.23) C6H5CH3 C6H5CH2 trapping

~

C-C6H1 H , C-C6H1 Radi ca 1 Constant current

c-CCHI1CH3 H,CH3,C-C6Hl1CH2, through mixture

'gH1 4 C6H139H hydrocarbon

'SH6

H,c-C H trapping glow discharge

C - C ~ H I 1 9 C-C6HqOCH3

C6H5 9 I'

c-c5H1 0 5 9

of argon with

d=I%;p=133 Pa; ( re f .4 , 3=1,2 W c m 12,151

Radical S i l e n t dischar- c4H1 0 C4Hg c5H1 2 (C2H5) 2CH trapping ge ( r e f . 24)

'gH1 4 'gH1 3 c-C6H1 2 c-C6H1 1 C7H16 C7H1 5

gH 5CH3 'GHgCH2

* ) Concentrations of any other r a d i c a l s a r e l e s s than 2% of ones shown i n the t a b l e ,

It was shown by s p i n trapping technique t h a t the main hydrocarbon r a d i c a l s generated during decomposition a re the products of breaking o f f hydrogen atom from r ings , chains o r methyl s u b s t i t u t i o n group (Table 5). I n the case of €IF

and MLV discharges i n benzeneproducts, t h e r fragmentation were observed ( r e f , 22). But this f a c t i s accounted f o r high s p e c i f i c discharge power (Fig. 1 ,2) followed by high gas temperature and high r a t e of secondary react ions. The l a t t e r a re responsible f o r the appearance of C2-C3-radicals i n t h e case of methane (Table 5). Quant i ta t ive da ta concerning r a d i c a l concentrations a r e

Mechanisms of decomposition of Hydrocarbons in electrical discharges 763

absent because l a t i v e concentration. Analyses of da ta of t a b l e s (3-5) show that among gaseous products a r e a few

t o confirm the conclusions of r e f , 7 about molecular break-ub of cyclohexane r ings. To c l e a r up t h i s question and t o study mechanism of chemical reac t ions spe- c i a l experiments were car r ied out with dc discharge by using isotope-labeled technique, The mixtures of protonated and deuterated cyclohexanes i n r a t i o s 9:l and 1:l with t o t a l concentration of 1% were introduced t o argon and used as working gas i n dc discharge. Products were analysed by chromat o- mass-spectrometry ( r e f , 4,15). The gaseous products were shown t o divide i n r e l a t i o n with i t s i s o t o p i c com- pos i t ion t o three groups. The first moue consisted of products, containing only hydrogen o r only deuterium atoms: ethene (C2H4;C2D4); acetylene(C2H2;C2D2); methylencyclopen-

cyclohexadienes (c-C6H8;c-C6D8); benzene(C6H6;C6D6). The second group consisted of products having comparable output a l l par- t i a l l y s u b s t i t u t e d proto-deutero compounds: ethane and propane (Table 6).

a l l known techniques a r e nonquantitative and give only re-

which can be generated from r a d i c a l s (Table 5). A t first view it could be

tane (C-c H -CH2;C-C D -CD2); cyclohexene (C-C~HIO;C-C~DlO); 1,3 and 1,4 5 8' 5 8'

TABLE 6. I so topic abundance of cyclohexane decomposition products included i n 2- d and 3- d group, see t e x t : &=I%; j=2,4 mA/cm2;

~ ~ 1 3 3 Pa. Sum of r e g i s t r a t e d i s o t o p i c abundance of every product is equal t o 100%. I so topic composition of i n i t i a l mixture C-c6H1 2: C-C6D1 2=B:l

Product Isotop Abundance,% Product Isotop Abundance,%

B=I B=9 B=l B=9

Ethane C2D6 12 5 Butene-I 18 7 (Butene-2 28 22 numbers

C2D5H C2D4H2 C2D3H3 42 66 i n bra-

ckets) Propane C3D8 12 I

C3D7H 13 4 C i D i H 2 9 7 C3D5H3 12 12 C;D&H~ 8 8

1 4 10 Methyl- C-C6DI1CD3 C3D3H5

C3DH7

C3D5H

C3D3H3

C3D2H6 I 5 22 CyClO- C-C6DqICD2H 17 36 hexane c-06Dq1CDH2

C-C6D1 CH3 Propene C3D6 51 32 c-C6H1 CD3

44 51 c-C6H,, CD2H 0 11 C-C6H1 CDH2 5 6 C-C6H1 1 CH3

C3D4H2

24 8 9

10 11 12 11 15

24 3 6

18 2 5 6

37

7 64 D. I. SLOVETSKY

The t h i r d group consisted of products having unequal isotopeoutputs:me- thylcyclohexane, propylene and butene (Table 6) . Direct primary decay of cyclohexane can give only products of the group. Re- l a t i v e output of various products of t h i s group depends on i n i t i a l hydrocarbon concentration i n mixture, current densi ty and gas pres- sure (Fig. 8 a ,b ,c ) .

I 433 266 p,Pe I 0

Fig. 8. Outputs of some stable products related to quantity of de- composited cyclohexane molecules as a functions of initial hydro- carbon concentration in mixture with argon (a), discharge current

1 -C2H4; 2 -C2H2; 3 -c-C6HI0(2,3)*); c-C6H8=CH2(23); c-C6H8-1 ,3 (76) i C-c6H8-1,4(60); c6H6(313) i c3H4(12); 4 -C2H6(1,5); C3H8(1,75) C H (1,8); 5 -c-C~H~~CH~(~O); C4H10(220); 6 -S4I%-1(14); C4H8-2 (20); cqHg-1,3(31); b) ~ ~ 1 3 3 Pa; d=1%; 1 'C2H4; c2H6(3,7); c3H8 (1295); CxHg(5); c4Hg-?(?6); c4%-2(12); 2 - C2H2; c3H4(14,2); c-C6HllCH;(6,5); c-C6HI0(3); c-C6H8=CH2(12); C6H6(10); C4H6-1 ,3; c) j=1,2 mA/cm2; A=?%; I - C2H4; C2H2(1,6); C H (25); 2 - C-C~H,~; C-cgHg(l4) 1 CgHg(10); c2H6(9) i 3 - c4H4-1,2(7) ; c4Hg-2(8,8) ; C4Hg- l,3; 4 - C5~=CH2(10); 5 - C-C6H4CH3 . * ) Factors by wich must be divided the figure values to obtain the true values a r e given in brackets

density (b) and gas pressure (c); a) p=133 Pa; j=1,2 mA/cm 2 ;

3 6

3 6

A number of them could be generated by means o f consecutive dehydrogenation of primary products during t h e i r c o l l i s i o n s w i t h hydrogen atoms and e lec t - rons. I n these cases t h e i r r e l a t i v e output could be increased by increasing

gen and electrons. But r e l a t i v e products outputs of first group were decrea- sed o r invar iab le as observed with current densi ty increases (Fig, 8b). T h i s f a c t excludes s i g n i f i c a n t contributionsgFconsecutive dehydrogenation react ions. The same is true about other reac t ions of dehydrogenation by col- l i s i o n s w i t h o ther r a d i c a l s a s generated i n primary decay. I n both ca- s e s r e l a t i v e outputs of theseproducts must be increased w i t h d . Indeed so- me increase was observed f o r C2H2, c-C6Hq0, c-C6H8 and C H -CH2 i n t h e range of small (Fig. 8a) and w i t h increase o f pressure (Fig. 8c) . But this dependence m u s t be more sharp i n t h e case of consecutive reac t ions , namely 4 - A and 4-p'. Weak dependence of r e l a t i v e outputs on d , p and current densi ty exclude e n t i r e l y contr ibut ion of consecutive dehydrogenation. Therefore a l l products of first group are d i r e c t products of primary decay of exci ted cyclohexane molecules by e l e c t r o n impacts (1,2). Dependence of

current densi ty because of increase of concentration of atomic hydro-

of output 5 8-

2

Mechanisms of decomposition of Hydrocarbons in electrical discharges 765

various f e r e n t path of exci ta t ion. By small d c o l l i s i o n s of cyclohexane with meta- s t a b l e argon atoms ( r e a c t i o n 2) give s i g n i f i c a n t contr ibut ion, but by l a r g e r

of e lec t rons give main contribution. Appearance poten- t i a l s of various products a r e d i f f e r e n t a s the decay takes place from vari- ous l e v e l s followed by v a r i a t i o n of products composition with e lec t - ron energy. Then primary decay of excited cyclohexane takesplace according t o the sche- m e : breaking of C-H bonds

product outputs on discharge parameters and 4 is accounted f o r d i f -

d ( L . h O , l % ) impacts

(71) (12) + a 2

C-c6H8 c-C6H6 + H2 .

Strong i so topic e f f e c t was observed f o r benzene, Its output due t o decay of c-C6DI2 was a t times higher than i n t h e case of c-C6H12. T h i s can be explained by s t rong influence of mass on the complex decay (12). Breaking of C-C bonds of types C2:C2:C2 :

(13)

C2H4 + C2H2 + C2H5 + H (15)

(16)

2C2H4 + C2H2 + H2 (14) E 3C2H4

( c-c6H1 2) *

and types C2:C4

and breaking off one C-C bond followed 'by isomerization of b i r a d i c a l and escaping of H2

(c-C6H12)* - C6HI2 - c-C5%=CH2 + H2 . (17) The f i rs t group of products includes possibly an a l len . Its i s o t o p i c com- pos i t ion w a s not studied. But depedence of a l l e n output on discharge parame- t e r s i s t h e same a s observed f o r products of first group (Fig. 8a,b,c) . Hen- ce it i s generated due t o reac t ion

(c-C6HI2)* - C4H6 + C2H4 + H2

(c-C6HI2)* - C3H4 + C3H7 + H (18) which i s only one example of r i n g d i s t r i b u t i o n of types C3:C3 with appearance of s t a b l e products, Decay of type C:C5 W a s not observed. Reaction scheme (9-18) d i f f e r s s i g n i f i c a n t l y from supposed e a r l i e r ( 7 ) , ac- cording t o which main decay reac t ions give sa tura ted alkanes d i r e c t l y (etha- ne, propane, butane). I so topic abundance of alkanes with near ly equal out- puts of a l l protonated-deuterated products as observed (Table 5) is the proof of with p a r t i c i p a t i o n of rad ica ls . Generation of r a d i c a l s can occur

generation of thesealkanes during secondary reac t ions most possibly

during primary decay of exci ted molecules:

c2H5 + c4H7 (19) C2H3 + c4Hs (20 ) E C3H7 + C3H5 . (21)

(c-c6H12)*

Direct experiments confirmed possibfe r a d i c a l s with simultaneous escaping of severa l hydrogen atoms and molecules and simultaneous breaking off severa l C-C and C-H bonds by d i r e c t impact o f e lec t rons w i t h molecules, For example,following products were observed as r e s u l t of butane dissociat ion: CH2(lJ,5'1); CH3(9,020,5); CH4(Il ,0+0,5); C2H2( 13,021,O) ; C2H3( 13,021 ,0) ; C2H5( 10,521 ,O) ; C3H3( 14,Ok?,O) ; C3H4( 14, O t 2 , O ) ; C31T4(12,5~1 , O ) ; C4H2(15,021 , O ) ; C4H3(14,0+1 , O ) ; C4H4(lLc,5-1 , O ) ; C4HS

appearance of s t a b l e products and

766 D. I. SLOVETSKY

(12,022,O); C4H6(12,021 ,O). Appearance potent ia ls as measured are shown i n brackets ( r e f . 25). The products of the th i rd group are generated as r e su l t of secondary reactions of primary decay products. Their unsubstituted pa r t s are products o f prima- r y decay. They are connected w i t h r ad ica ls , generated i n primary decay and/ o r i n secondary reactions, Methylcyclohexane is one of most in te res t ing productsof this group. It has en t i r e ly protonated o r en t i r e ly and output of products with a l l pa r t ly deuterated and protonated methyl groups are nearly nation of radicals

o r due t o including of methyleneradicals t o C-H - bond

deuterated r ing connected with methyl group

equal (Table 6). Hence it can be generated by two paths:recombi-

(22)

C-C6H12 + CH2 - C-c6H11c~13 , ( 2 3 )

M , walls C-C6Hll 4- cl13 '-'GH1 1 CH3

I n the l a t t e r Case C-C6Hq1CH2D; C-C6D11CH$ and C-C6Dl,,CD2H must not be ob- served. It is contrary no contribution and generation i s due t o recombination of radicals (22). Me- thyl rad ica ls must be nearly equally deuterated - protonated. Such radicals can explain generation of ethane - c2% (24)

t o dgta (Table 6) . Hence the react ion (23) gives

hl, walls CH3 + CH3 and propane also ~

w i t h t h e i r ' c3%

CH3 + C2H5 M , walls

isotopic composition (Table 6 ) . Butenes and propylene are generated most probably due t o disproportion reac- t ions C H + CH3 - C4H8 + H2 , (26) 3 7

C . H 2 5 + CH3 - C3H6 + H2 (27) and radical recombination

M , walls C4H7 + H a c4H8 M,walls C3H5 + H C3H6

Thesestable-products of r ing breaking are C2:C2:C2 type (C2H4, C2H2), one C-C bond (C5%=CH2) and breaking of several C-H bonds (C6HI0, C6H8, C6H6) are generated during primary decay of excited cyclohexane molecules. The ring breaking of C4:C2 (C4H6) and C3:C3 (C3H4) typesmakes a small contribution. Total output of gaseous s tab le products due t o d i r ec t breaking of C-C bonds i s l e s s than 7,576 and of C-H bonds i s l e s s 2,576. Output of gaseous products generated due t o secondary reaction, w i t h par t ic ipat ion o f cyclohexyl radi- ca l (main rad ica l i n plasma) namely methylcyclohexane i s l e s s tha t 0,476. Outputs of gaseous products of recombination of other radicals are one order value la rger i n sp i t e of the other rad ica ls concentrations as measured are a t 5 times l e s s than cyclohexyl one. This allowsto suppose tha t cyclohexyl rad ica ls take pa r t i n building o f so l id products on the walls.

MECHANISMS OF DEPOSITION OF FILMS

The so l id polymeric f i lm is the main product of decomposition of a l l studi- ed hydrocarbons i n dc discharge (Table 3). Rate o f film deposition increases proportionally t o cyclohexane concentration i n i n i t i a l mixtures under con- d i t i on of constancy of a l l other plasma parameters including the charged par- t i c l e flow t o surface. Hence deposit ion does not occur due t o put i n t o f i lm of cyclohexane ionsbeing the main ions i n plasma at dr 0,1%. To study con-

Mechanisms of decomposition of Hydrocarbons in electrical discharges 767

t r i b u t i o n i n deposi t ion of unsaturated hydrocarbons generated due t o hydro- carbon decomposition a numbwqexperiments were c a r r i e d o d - the mixtures of deuterated cyclohexane and protonated ethenewere used by t o t a l concentration of hydrocarbons i n argon equals 1%. Deposited f i lms do not con- t a i n t h e hydrogen i f etheneaddition was equal t o one generated due t o decom- pos i t ion of cyclohexane i n mixture with argon (Table 3 ) . The hydrogen was observed i n films only by increasing tfte ethene concentration i n mixture t e n times (3% C2H4 +I% c-C6Dq2 +96% A r ) ( r e f , 4,5,15). Hence the unsaturated i n plasma do not give s i g n i f i c a n t contr ibut ion t o f i lm deposit ion. The most probable mechanism of f i l m deposi t ion as s tudied contains building i n of heavy r a d i c a l s generated due t o breaking of C-H bonds. B u i l d i a up i s due t o recombination w i t h f r e e bonds on surface generated due t o i t s bombard- ment by energet ic plasma p a r t i c l e s . The l a t t e r givesriseto f u t h e r des t ruc t ion of bonded r a d i c a l s , atomic hydrogen and l i g h t r a d i c a l desorption (CH2, CH3 ...). Free bonds generatedasa r e s u l t recombine one w i t h another giving gene- r a t i o n of cross links and double bonds, and w i t h r a d i c a l s coming from plasma giving films growth. Recombination w i t h hydrogen atoms leads t o decay of f r e e bonds, T h i s reac- t i o n s explain mixed s e Such a mechanism is confirmed by da ta observed using benzene.In t h i s case film contained aromatic s t r u c t u r e s - phenyl, bi- and three- phenyl fragments. Concentrations of the l a t t e r were diminished w i t h increasing bombardment i n t e n s i t y by means of decreasing of benzene concentration i n i n i t i a l mixture, or addi t iona l treatment growed f i l m s by pure argon plasma. Main r o l e was shown i n f r e e bonds generation t o play ions by using mixture of methane w i t h r a r e gasek. Ion building i n f i lms is not s i g n i f i c a n t as s u b s t i t u t i o n of hydrocarbon ions by xenon ones do not influencethe r a t e of f i l m growth ( r e f . 13,26). Hence growth i s due t o hydrocarbon rad ica ls . Cyc- lohexyl r a d i c a l s primary decomposition and give

t o

i so topic composition of l i g h t r a d i c a l s i n plasma, becau- hydrogen is main gaseous product of hydrocarbons decomposition.

b u i l t i n t o f i lm i n cyclohexane being main products o f r i s e t o very low outputs of gaseous products,

CONCLUSION

This appl ica t ion of quant i ta t ive method t o studying of chemical r e a c t i m uechanism i n nonequilibrium plasma diagnost ic ( r e f . I) allowsto f ind m o s t probable paths t i o n of hydrocarbons i n dc di'scharge plasma ( r e f . 4,5,15). The mechanism d i f f e r s s i g n i f i c a n t l y from those Pvoposed e a r h v ( r e f . 7,21,22 e t c ) . The waLn dttfrremce A h o w n % w e i s t h a t primary decay g ivesr i se t o r a d i c a l s even i n the react ions generating s t a b l e products too. Most decays a r e due t o breakybof C-H bonds followed by s o l i d f i l m deposi t ion a s r e s u l t of ra- d i c a l recombination on surfaces . Breaking of C-C bonds gives riseto gaseous products. Primary decomposition reac t ions a re decays of e l e c t r o n i c a l l y exci- ted hydrocarbon molecules, Exci ta t ion i s made by c o l l i s i o n s w i t h e lec t rons and meto,- s t a b l e atoms. Role o f ions i n decomposition reac t ions is negl igible . They play la rge r o l e i n s o l i d product deposi t ion r a t e and composition due t o bombardment of sur- faces .

of gaseous and s o l i d products generation during decomposi-

768 D. I. SLOVETSKY

The mechanism is valid in glow discharges by low pressures ( p z 500 Pa) and current densities ( d< 10 W c m ). Under other conditions f o r example in discharges with higher specific poer and pressure particularly in HF and Mw discharges a new detailed analyses is needed based on the quantitative diag- nostics of plasma.

2

REFERENCES

1. D.I.Slovetsky, Chemical reaction mechanisms under nonequilibrium plasma

2. L.S.Polak, Nonequilibriume chemical kinetics and its application, Moskow,

3. Ju.A.Ivanov, I.N.Karabashev, L.S.Polak, High enerm chemistry, 2, 75-79,

4. G.A.Alekperov, L.Y.Polak, D.I.Slovetsky. In: Syntheses in low temperatu-

5. G.A.Alekperov, L.S.Polak, D.I.Slovetsky, Hinh energy chemistry, 'M, N6,

conditions, Moskow, Science, 3 l l p . (1980) (in Russian).

Science, 248p. (1979) (in Russian).

(1979) (in Russian; the journal is translated in English).

re plasma, Moskow, Science, 44-46 (1980) (in Russian).

457 (1977); 12, 379-38 (1978); 12, N1, 85-90 (1979) (in Russian; the journal is translated in English).

6. Ju.A. Ivanov, I .V.Soldatova, 1.L.Epshtaine , ibid, 3, N5, 465-470 (I 985). 7. H.SchGler, G.Arnold, Z.Naturforsch, B, 670-675 (1962); a, 604-607

8. S.Klagge, Beitr.PlasmaDhys., 3, 309-317 (1975). 9. R.Winkler, J.Wilhelm, ibid, 8, 403-422 (1968); 11, 159-178 (1971).

(1963); z, 435-Ml (1965).

10. V.E.Galtsev, Ju.A.Ivanov, N.M.Ritova et all. Proc. 6-th Allunion Confe- rence on Physics of low temperature Plasma, Leningrad, USSR, 1, 132-134 (1983) (in Russian).

11. M.Deutsch, H.Sabadi1, E.W.G'drse, BeitrGe aus der Plasmaphys., Q, 183-

12. G.A.Alekperov, V.E.Zubarev, L.S.Polak, D.I.Slovetsky , High energy chemi- strg, 2, N2, 188-192 (1979) (in Russian; the journal is translated in English).

13. Ju.A.Ivanov, 1.V.Soldatova. In: Physical and Chemical Processes in low temperature plasma, Moskow, Inst. of Petrochem. Syntheses Academy of Sciences of USSR, 5-54 (1985).

190 (1975).

14. L.G.Piper, J.E.Velazco, D.W.Setser, J.Chem.Fhys., 2, N 6 , 3323-3340

15. D.I.Slovetsky. In: Chemistry of Plasma, Moskow, Atompress, N8, 189-229 (1981) (in Russian).

16. D.I.Slovetsky. Ibid, N I , 190-193 (1974). 1nt.J.Radiat. Phys.Chem., 8 , 257-282 (Pergamon Press, Great Britain).

17. P.Capezutto, F.Cramarossa, R.D'Agostino et all, Beitr. Plasmaphys., z, 18. D.I.Slovetsky, P.A.Ssrgeev, Chem.Review, w), 231 (1983). 19. G.Herzberg, The Spectra and Structures of Simple free Radicals, Cornell

20. G.Smolinsky, M.J.Vasile, Int.J.Mass.Swctrom.Ion.Phys., l8, 179-192 (1975). 21. H.Sch:ler, L.Reinbeck, Spectrochemica Acta, 6 , 288-301 (1954). 22. M,Durval, A.Theoret, J.ADP~. Polymer Sci., x, 527-537 (1973). 23. S.Kikkawa, N.Nomura, S.Morita et all, Chem.Lett. (Chem.Soc.JaDan), lo,

24. D.B.Hibbert, J.B.Robertson, M.J.Perkins, J.Chem.Soc.Faraday Transact. 1,

25. A.Nichaus, Z.Naturforsch,, G, 690-700 (1967).

(1973) -

205-220 (1977).

Univ.Press., Ithaca, Lond. (1971).

1337-1340 (1974)

22, 7499-1508 (1971)